During the past year, this laboratory continued studies on the mechanism of HIV entry, and on the development of novel protective and treatment strategies based on the molecules involved in entry. 1) Molecular mechanisms of HIV Env-mediated entry/fusion. a) Efficiency of fusion/entry measured by different assays. We have used both a vaccinia-based cell fusion assay and HIV infectivity assays (MAGI cells, or PBMC) to study various aspects of Env function. By several criteria, the cell fusion assay appears much more senstive. First, higher concentrations of blocking agents (antibodies, drugs) are needed for fusion compared to infectivity. Second, fusion between Env-expressing cells and targets cells expressing coreceptor but no CD4 can be induced by soluble CD4 (sCD4) to give a potent fusion response; by contrast, virion entry into such targets is induced much less efficiently by sCD4. Third, complementation by Env subunits containing different fusion inactivation mutations is readily detected in the cell fusion assay, but not in the virion entry assay. Thus the cell fusion assay has a major advantage for detecting relatively inefficient fusion events. However, it is more difficult for detecting blocking by inhibitory agents, and may over-estimate the potency of inefficient fusion events. The results have important implications for the mechanism of HIV entry, and the development of entry inhibitors, b) Mechanisms of antibody resistance of some primary HIV-1 strains. We have continued to develop assay systems to distinguish whether antibody resistance of some HIV-1 strains is due to absence vs. masking of the corresponding epitopes. The results will have important implication for the development of anti-HIV vaccines based on neutralizing antibody. 2) Novel anti-HIV agents based on HIV Env/receptor interactions. a) sCD4-17b, a potent HIV-1 neutralizing agent. sCD4-17b is a recombinant chimeric protein containing soluble CD4 attached via a long flexible polypeptide linker to a single chain antibody that binds to a CD4-induced epitope of gp120 involved in binding to coreceptor. While the protein neutralizes some HIV-1 strains with very high potency, other strains appear resistant, despite the known presence of binding sites for CD4 and 17b. Efforts are underway to solve this problem by producing variants of sCD4-17b with longer linkers. Furthermore, progress has been made in engineering Lactobacillus to produce sCD4, both in secreted form (to neutralize free virus) and in cell-anchored form (to trap virions). The goal is to develop a method of protection against sexual transmission of HIV-1 that is durable, economically feasible, and controlled by women. b) 3b3-PE38, an anti-HIV immunotoxin. Our previous studies indicated that this immunotoxin is significantly more potent in vitro and less hepatotoxin than a previous agent sCD4-PE40. Collaborative efforts during the past year have led to the design of even more potent variants, which also display greater stability. Collaborative studies using a SCID-hu mouse model indicate that latently infected T cells can be induced to activate HIV expression, thereby becoming sensitive to immunotoxin killing. We continue to focus on developing 3b3-PE38 for Phase I clinical trials in individuals whose virus loads have been effectively suppressed by HAART.